serum and milk in grey seal
mothers and pups during
lactation
ABSTRACT
Grey seal mother pup pairs of two UK colonies were sampled for blubber, serum and milk at three different stages of lactation during three breeding seasons (2004-2006) and samples were analysed for fatty acid (FA) profiles. This study provides information on how the dietary differences between populations are reflected in serum, milk and ultimately, in the offspring blubber. We also
investigate the differences between compartments within and between lactation stages to give an insight into the changes in FA composition during lactation, but also to determine the milk FA in relation to the blubber FAs and if pups blubber resemble that of their diet. Samples of serum and milk were collected at three lactations stages; blubber from mothers was collected at early and late lactation, while pup blubber was only sampled at late lactation. The results show that blubber, serum, milk and pups blubber could be differentiated from each other during the three lactation stages. The results show that the dietary differences between populations observed in the blubber FAs at early lactation were also seen in other tissues and in pups blubber. Although the dietary differences were
transferred to other tissues maternal blubber could still be well distinguished from the FA composition of the milk and from the serum during the different lactation stages. The FAs composition of the milk changed gradually across lactation stages. These changes were similar in both colonies and between years. The FA
composition of the pups at late lactation differed from that of their diet and that from their mothers‟ blubber. Although, the overall FA composition of pup blubber did not differ greatly between colonies, they could still discriminate colonies indicating that proportions of certain FAs that distinguished the mothers are probably transferred via the milk. This study provides important information on the ongoing debate of the possibility of use of milk to infer diet. It also provides a better understanding on the FA dynamics during lactation.
INTRODUCTION
In order to understand the interactions of marine mammals with the marine ecosystems of which they are part, detailed information about the diet of these animals is essential. Determining the diet of marine mammals is challenging, and a variety of approaches such as stomach lavage and faecal analysis have been developed (Pierce & Boyle 1991). Although these have proven to be valuable tools they are also subject to several limitations and potential biases (Jobling & Breiby 1986, Pierce et al. 1993) . For instance, the information obtained from these types of analyses will mainly represent the most recently ingested prey, or may be biased towards prey species that contain hard parts that are less easily digested. Alternative methods such as stable isotope ratios (Gannes et al. 1997) and fatty acid signature analysis ( Iverson et al. 1993, Iverson 1993) have demonstrated their usefulness for overcoming some of the problems inherent in faecal or lavage sampling techniques.
The basis for using fatty acids (FAs) as a dietary indicator is that dietary FAs are assumed to be accumulated into the blubber of marine mammals with little or no modification (Iverson 1993, Iverson et al. 1997, Smith et al. 1997) and the analysis of FA composition of blubber has therefore become a popular method for providing information about the accumulated dietary history of the previous fattening period (Iverson 1993, Iverson et al. 1995, Iverson et al. 1997). Several studies at different trophic levels in terrestrial and marine ecosystems have shown that the composition of dietary FAs are reflected strongly in the composition of adipose tissue (Beck et al. 1994, Pond et al. 1995, Castell et al. 1995, Iverson et al. 1997, Smith et al. 1997) and the general approach is now widely accepted despite disagreements over the accuracy of the estimated dietary composition (Beck et al. 1994, Pond et al. 1995, Castell et al. 1995, Iverson et al. 1997, Smith et al. 1997, Grahl-Nielsen 1999). However, there is still considerable debate among
researchers as to how the FAs are deposited, transfer and to what extent the FA composition of various body tissues reflects the diet. To determine to what extent
with that of a controlled diet (Tollit et al. 2006, Nordstrom et al. 2008). These studies are generally based on only a few individuals and do not represent the complexity of free-ranging animals. Lactating mothers with their pups have been also used to investigate the influence of the diet in the blubber FAs (Ackman and Jangaard 1965, Iverson et al 1995, Grahl-Nielsen et al 2000, Birkeland et al 2005). In hooded seals (Cystophora cristata) it was found that the blubber of the pups was very similar to their milk diet (Iverson et al 1995). However, Grahl-Nielsen et al (2000) found that the blubber FA composition of grey seals (Halichoerus
grypus) pups was more similar to their mothers blubber than to the milk. Many of these studies describing the deposition and transfer of FAs has been based in a small sample size, some of them are cross-sectional studies and others only take into account samples from an early lactation stage.
Diet inference has also been done using FAs in other lipid-rich body tissues. For instance, Iverson (1993) suggested that the diet of female marine mammals could be inferred from the FA composition of their milk. This is predicted on the assumption that lipids secreted into the milk by a fasting female will be a
reflection of the lipids from the blubber and therefore the diet. During fasting the FAs accumulated in the adipose tissue are released into the circulatory system and will be circulated to other organs and tissues, including the mammary gland. Therefore in fasting seals the FA composition of milk will be almost exclusively determined by the FA release from blubber, with very little de-novo synthesis occurring in the mammary gland (Iverson 1993). In contrast, the milk FA
composition of animals that feed regularly during lactation (such as otariids) will be strongly influenced by FAs from recent dietary intake, since circulating lipids will be directed to the mammary gland (Iverson et al. 1997). If dietary intake is insufficient for milk production, FAs will be mobilized from adipose tissue, and the FA composition of the milk will be a reflection of a combination of both direct dietary and stored blubber lipids (Iverson 1993, Iverson et al. 1995, Iverson et al. 1997). The effect of diet on milk composition has been studied in dairy breeds, humans and domestic animals, but there are few studies that discuss the
implications of the use of milk as a dietary indicator for marine mammals (Iverson 1993, Iverson et al. 1995, Iverson et al. 1997, Grahl-Nielsen et al. 2000, Staniland & Pond 2005).
The purpose of this study was to determine the way in which accumulated dietary FA reserves by females are transferred to their offspring via milk and the deposition of this in the offspring blubber. This study also intends to observe to what extent any dietary signature in female blubber is retained in the blood and milk of the female and how the diet of the pups (milk) is reflected in the
accumulated FAs of their blubber with the aim to observe if it is possible to use milk as a dietary indicator. Female grey seals give birth to one pup each season, and they remain with their pups throughout the ~18-day lactation period, relying entirely on stored reserves (i.e. blubber lipid). By sampling the blubber, blood and milk of known mothers and their pups at different times throughout lactation we can examine 1) the differences in FA composition of different body compartments at specific time points, 2) the changes within compartments throughout lactation and 3) differences between the FA composition of the pup‟s blubber and that of the mother‟s milk and blubber. By studying two geographically separated populations, which have previously been shown to have distinct blubber compositions, presumably reflecting differences in diet (Walton et al. 2000, Walton & Pomeroy 2003, Arriola et al. submitted), we can also examine whether differences in female blubber at early lactation are maintained throughout the different compartments of the female and in the offspring. The FAs that are released during the fast to the bloodstream will have a direct effect on the FA composition of the milk and thus in the FA composition of the pups. We expect to find similarities between the FAs from blubber and milk if the FAs are transferred with no or little modification from one compartment to another. Since the
mobilization of FAs from blubber is selective, the variation in FA composition in milk, blood and blubber across lactations stages should be very similar between populations.
METHODS
PO PULATIO N A ND SAM PLE CO LLECTIO N
Grey seal mother-pup pairs were studied at North Rona (NR) (59o06‟N, 05 o50‟W) and Isle of May (IOM) (56o10‟N, 2 o33‟W) during three consecutive breeding seasons (2004-2006). The total numbers of females and pup sampled for each year are shown in the FA profiles summary tables, Appendices 2-7. The average lactation duration was 18 days. In 2004, mothers and pups were captured at two lactation stages: early (the first quarter of total duration of lactation) and late (the last quarter of the duration of lactation) with a minimum interval period of 10 to 12 days between them. At both captures females were sampled for blubber, blood and milk. Pups were sampled for blood at both captures and only blubber at late lactation. In 2005 and 2006 females were captured three times at intervals of 6 to 8 days (early, mid and late lactation stages). Females were sampled for blood and milk at every capture and blubber only at early and late lactation. Pup blood was sampled at each of the 3 stages while pup blubber was sampled only at late lactation. Females were immobilized using an intramuscular doze of „Zoletil 100‟ (Virbac, Cedex). A small incision on the mid pelvic region was made with a scalpel and a full depth blubber core was obtained using a 6mm diameter punch
(Acuderm Inc). Since FA composition can vary across the body depending where the animal is sampled all the blubbers were collected from the mid-pelvic region (Walton et al. 2000). The samples were stored in chloroform:methanol (2:1) with 0.05% of BHT as antioxidant. At each capture blood samples were drawn from the extradural vein using the Vacutainer (Becton Dickinson, UK). At the end of each day, samples were centrifuged and aliquots of serum and plasma were
transferred to microtubes and stored at –20oC until further analysis. To stimulate the release of milk 1 mL of oxytocin was intravenously injected and 30 mL of milk were obtained. Samples were collected in a hexane-washed glass jar and stored at –20°C until further analysis. Sample sites were treated with topical antibiotic and a mass specific prophylactic intramuscular injection of antibiotic was
drawn from the hind flipper or extradural vein. Pup blubber samples were obtained following the same procedure as in females but using a 4 mm diameter punch (Acuderm Inc), following an IV Zoletil dose equivalent to 50% of an adult mass based dose. Body composition was estimated as described previously in Chapter 3. Morphometric measurements: adult mass, girth and length and pup mass were recorded at each capture. All procedures were performed under UK Home Office Licence.
FATTY AC ID ANALY SES Blubber
Blubber lipid was extracted following the method described by (Folch et al. 1957). Samples were homogenized with 10 ml of dichloromethane:methanol (2:1
vol/vol). The top surface was washed by adding chloroform:methanol:H2O (3:48:47 vol/vol) and treated with anhydrous Na2SO4 and dried under nitrogen, weighed and dissolved in toluene at a concentration of 100mg/ml. Fatty acids were esterified to produce fatty acid methyl-esters (FAME) using 1% (vol/vol) of sulphuric acid in methanol. Samples were then incubated overnight at 50oC. The purified FAMEs were dissolved in hexane. FAMEs were analysed by gas
chromatography using a Trace GC-2000 (Thermoquest, CE Instruments) equipped with a flame ionisation detector and fitted with a DB23 fused silica capillary column (25 x 0.25mm, J&W Scientific). Hydrogen was employed as the carrier gas. The temperature was programmed to start at 60oC for 2 min, then rise to 150 oC at 20 oC min-1, then held for 2 min before a further rise to 205 oC at 1.8 oC min-1and finally rising to 230 oC at 5 oC min-1. Separated components were
identified by reference to standards (Chapter 1). Milk
Dichloromethane was added to aproximately 0.5 - 1.0 g of milk sample. After a few minutes it was strained with fibre and dichloromethane:methanol 2:1 (vol) was
dichloroform:methanol:water (2:50:50 vol/vol). Sodium sulphate was added and dried under nitrogen, weighed and dissolved with toluene at a concentration of 100mg/ml. The fatty acids were esterified and analysed by GC as previously described in the analysis of blubber samples.
Serum
To a previously weighed 2 g of serum 10 ml of dichlormethan:methanol 2:1 (Vol) was added followed by 50 μl of C:23. The final mixture was strained with
fiberglass wool, and 30 extra ml of dichloromethane/methanol were added. Salt solution (1%) was added and left for 1 hr. After this stage the same procedures as in milk were followed.
STA TISTICA L ANALYSIS
Fatty acid concentrations are reported as percentages of total FA. These relative values were arcsine square root transformed prior to statistical analysis. Samples of blubber, serum and milk were divided into three lactation stage categories based on the pup‟s age when a sample was collected relative to the observed birth and weaning dates. Because lactation duration varied between females, samples collected during the first quarter of the total duration of lactation was designated as early, the last quarter as late and samples collected during the intervening period were designated as mid lactation.
Random forest analysis (see Chapter 2) was used to determine to what extent the compartments (blubber, blood and milk) could be distinguished by their FA profiles and how the FAs vary between lactation stages within a given
compartment. This gave a class variable with 8 classes (blubber samples were not collected during the mid-lactation capture). These analyses were performed for all the animals that had a minimum of paired samples for two compartments
(blubber-blood, blubber-milk, blood-milk). We also compared the FA composition of mothers with that of pup blood and blubber within the same analysis.
To further examine the degree to which each FA contributed to differences between compartments, lactation stages and colonies observed in the random
forest analysis, multiple pair wise comparisons tests were carried out according to the methods described more fully in Chapter 3. Here we focused our attention to 1) within compartment changes of individual FAs throughout lactation and 2) colony differences of individual FAs in specific body compartments at specific times through lactation.
RESULTS
A total of 630 samples (blubber, blood and milk combined) from mothers and pups were analyzed during the three years of study. Total numbers of females sampled each year are shown in the summary tables, Appendices 2-7, which also present the relative percentages of each of the 60 FAs identified in the three compartments for mothers and pups for both populations for each year. During each of the 3 years the most abundant FAs in all three compartments and in both colonies were the saturated FAs (SFAs) 14:0, 16:0 and 18:0, the monounsaturated (MUFAs) 16:1n-7, 18:1n-7, 18:1n-9, 18:1n-11, 20:1n-9 and 22:1n-11 and the
polyunsaturated FAs (PUFAs) 18:2n-6, 18:4n-3, 20:5n-3, 22:5n-3 and 22:6n-3. Overall, MUFAs comprised the highest proportions of the total FAs in all body compartments. In both populations SFAs were more abundant in milk than in blubber, the total MUFAs were higher in blubber while PUFAs comprised were similar proportions in all compartments.
CLA SSIFICA TIO N U SING RA ND OM FO REST
Random forest analysis showed a clear distinction in FA composition between the two colonies, between the three compartments of females, the two compartments in pups and also between mothers and pups (Appendix 8 A and B). NR and IOM were clearly discriminated by the FA composition of all the body compartments.
compartment. The major misclassifications occurred between lactation stages within a compartment, while a small number of samples were misclassified into the same compartment but in the wrong year (11.6%) or the wrong colony (6.8%). Only eight of the 630 samples (i.e. ~1.3%) were incorrectly classified into a
different body compartment.
Among IOM samples (Appendix 8A), blubber from 2004 showed major misclassifications between lactation stages with none of the early samples classified correctly. There was also a poor distinction between mothers and offspring. There was greater discrimination among blubber samples in 2005 and 2006, with only a few samples allocated to the wrong year and colony. Milk samples from IOM were well distinguished from the other compartments and between lactation stages. The samples that were misclassified into an incorrect year and/or colony were nevertheless allocated to the correct compartment (i.e. milk) and lactation stage. Blood samples were well distinguished from the other two compartments with not a single sample being misclassified into a different compartment. However, there was relatively poor distinction between lactation stages, and several samples were wrongly allocated to NR blood samples.
The pattern among samples from North Rona was broadly similar to that at the IOM (Appendix 8B). Misclassifications among NR blubber samples in 2004, 2005 and 2006 occurred mainly between lactation stages. Blubber samples from mothers were well discriminated from pup blubber. Milk samples were well discriminated from the other two compartments, with only 1 of 87 samples classified as blubber. Milk samples from 2005 had the poorest classification among milk samples with 27 out of 31 misclassified. In 2006, the
misclassifications were relatively low in comparison to the previous years. Blood samples were clearly differentiated from the other two compartments and the misclassifications occurred mainly between the mid and late lactation stages. Mothers and pups were also clearly differentiated. Within pups, blubber samples were relatively well classified between lactation stages.
The FAs within and between compartments varied between lactation stages. At early lactation differences between blubber and milk could be seen in over 20 FAs in both populations for all the years of study to the exception of IOM 2004 where the compartments only differed in 10 FAs. In both colonies, FAs found in
high proportion such as 16:0, 18:0 and 14:1n-5 significantly differed between blubber and milk at early lactation. In addition FAs 16:1n-7 and 20:1n-9 were only different in NR samples. The first three FAs mentioned are part of the most
important variables selected by the random forest for splitting between the groups.
CHA NGES IN FA C O MPO SITIO N THRO U GHO U T LAC TA TION
Changes in FA composition of the blubber of mothers were described in Chapter 3. Briefly, we found that FAs are selectively mobilized resulting in substantial changes in the blubber FA composition over lactation. Here we focus on maternal blood and milk and on pup blubber.
Overall, the FAs composition in maternal serum remained relatively constant throughout lactation in both colonies (Fig. 1). Only a small number of FAs (0-3 out of 60) changed significantly (Table 1). In contrast, the milk FA composition
changed gradually throughout lactation (Fig. 2). While few FAs changed
significantly between consecutive stages (2-6 out of 60) the number of FAs that changed significantly from early to late lactation was much greater (24-28 out of